Creepy vs Useful - getting personalisation right
Mention personalisation and most people have an idea of what it means. The responses tend to range from “creepy” to “useful” but the reality lands somewhere between “relevant”, “helpful”, and "gimmicky" (also referred to as "pointless personalisation").
In fact, the proof is in the performance of personalised video – where, “relevance” and “helpfulness” translates across the board into deeper engagement and more effective conversions from the video. Viewers are more likely to view the video to the end (86% view through rate on average), and click through from the video (280% average uplift in CTR compared with a generic video).
So what does data actually look like when applied correctly?
Using browsing data for retargeting and dynamically-generally creative has a reputation of being a bit of a clumsy, relentless stalker that shadows your every move. The pair of shoes that follows you around on every web page you visit for weeks after you’ve bought them doesn’t do much for the reputation of data-driven marketing.
Personalization at scale is critically important because you and I - as individuals would much rather receive a message which was tailored to my need
– Sir Martin Sorrell, Executive Chairman at S4 Capital
But what if that data is put to good use to help reduce wastage in the creative?
What if we can do away with the one-size-fits-all approach to video marketing, and use the data to focus on elements that may specifically be of interest to the user?
The idea that you would treat your customers as strangers makes no sense
– Simon Andrews, Addictive
Take this campaign for the Fiat 500X with Auto Trader. In this instance, we’re talking about a vehicle that has many functions and features that have different degrees of interest and relevance to different audiences. If a video has a very limited amount of time to grab the attention of the viewer (and in Instagram Stories this is on average just 7 seconds), we can't show everything – especially if some of those features won’t be of interest to the viewer.
For instance, if a user is interested in dance music and festivals they may want to see more of the sound system, or if they’re interested in camping or dog walking, then the boot space is probably going to be their most interesting feature. The highest performing interest/demographic of this campaign was Fishing, Male and Dog Walking - now there's another follow-up task looking at optimising Dog Walking Fishermen, but that's another story.
By using dynamic video, the campaign was able to generate a vast array of different personalised Instagram Stories video creatives focussing on the things that resonated most with the viewer using their Facebook Interests to steer the content. And the results confirmed the theory – with higher performance of niche interests such as fishing and camping.
This approach to crafting an interest-based campaign is just the beginning of a journey. The campaign was also filtered by Auto Trader user data, focussing solely on an audience of users who had actively searched for a new car in the past 90 days, then once the user had seen the initial “teaser” content as an Instagram story, we were able to deliver a more specific product-focused creative, as well as any offers from their local dealership – moving the customer down the funnel and closer to a test drive or purchase.
You can read more about the Fiat 500X campaign here